To create this transcript, I split the episode into 15-minute chunks with Fission, because Micro.blog’s OpenAI-based transcription is limited to 15 minutes. Then uploaded the MP3s to Micro.blog and copy/pasted the resulting transcripts into BBEdit. Fixed mistakes and edited to be more readable.
Manton Reece @joshua No, I’ve tried Aiko on Mac which I assume is similar. But it is very, very slow on my Mac. Running it through OpenAI only takes about 30 seconds.
Josh Withers apparently it is based on the same tech but MacWhisper uses a few extra CoreML goodies. But I was surprised at how not-quick it was on my pretty well spec’d MBP M2 Max.